Muscular Dystrophies and Allied Disorders V : Facioscapulohumeral, Myotonic and Oculopharyngeal Muscular Dystrophies

Facioscapulohumeral Muscular Dystrophy

Facioscapulohumeral muscular dystrophy (FSHD) is one of the most common muscular dystrophies, with an estimated prevalence of 1: 8300 to 1:15,000 and a high frequency of sporadic cases and variable penetrance ( , ). Penetrance, based on clinical presentation, is age-dependent, and most cases show clinical signs before the age of 20 years old, but onset can be at any age, including childhood ( , ). Anticipation is also well recognized with earlier onset in successive generations. Severity may relate to size of the deletion of the fragment associated with FSHD.

The disease was linked to chromosome 4q35 many years ago, but it is only relatively recently that the complex molecular genetics have been unravelled and two forms, FSHD1 and FSHD2, identified ( , , , , ). The molecular defect in the majority of cases (FSHD1) is a deletion of copies of a 3.3 kb DNA repeat fragment in the subtelomeric region of chromosome 4q (D4Z4), in association with a 4qA161 haplotype. In normal individuals the fragment varies in size from more than 50 to 300 kb, with 11–100 or more D4Z4 repeats, but FSHD1 patients only have 1–10 repeats, ( , ). Each D4Z4 unit contains a copy of the DUX4 gene, which is located next to a regulatory region of DNA on chromosome 4 known as a pLAM sequence. This is required for the function and production of DUX4 protein. The functional form of pLAM is described as 4qA or ‘permissive’, whereas the non-functional sequence is known as 4qB or ‘non-permissive’. The DUX4 gene is silenced in adult tissues, but in FSHD1 loss of D4Z4 repeats in association with a permissive 4qA haplotype, in particular 4qA161 and the rare variants 4qA159 and 4qA168, results in hypomethylation and ‘chromatin relaxation’. This allows re-expression of DUX4 protein, which is toxic in adult muscle. This molecular mechanism occurs in about 95% of patients with an FSHD phenotype and causes FSHD1. In the other 5% of phenotypically similar patients there is a low, but normal, number of D4Z4 repeats (FSHD2). In FSHD2 the repeats are hypomethylated on both alleles, compared with only one allele in FSHD1, and also the similar D4Z4 repeats on chromosome 10. In most FSHD2 patients the widespread hypomethylation is caused by a deletion in the SMCHD1 gene on chromosome 18. The protein product of SMCHD1 binds to D4Z4, influencing methylation and resulting in expression of DUX 4. This only occurs in the presence of a 4qA permissive haplotype with a polyadenylation signal that stabilizes the DUX4 transcript. Variants in another gene, DNMT3B, have also been identified in rare cases with FSHD2 ( ). Thus, both FSHD 1 and FSHD2 result from re-expression of DUX4, by different mechanisms, in association with a permissive 4qA haplotype. Studies suggest that mutations in the SMCHD1 gene may act as a genetic modifier that influences severity of the disease in people with FSHD1 ( ).

The diagnosis of FSHD1 by molecular analysis is now highly reliable, provided a double digest with EcoRI and BlnI restriction enzymes is carried out to distinguish the chromosome 4q fragment from the similar fragment on chromosome 10q26, both of which are detectable with the p13E-11 probe. The chromosome 10 fragment contains a BlnI restriction site that is absent from the chromosome 4 fragment. Thus, these two enzymes completely digest the chromosome 10 fragment, leaving the chromosome 4-related fragments. Confusion can arise, however, as interchromosomal exchange of the repeat regions occurs in a few individuals in the normal population, resulting in hybrids of chromosome 4 and 10 fragments. Germ-line mosaicism may also hamper molecular analysis, and false-negative and false-positive results can be obtained and require the use of additional probes. The double-digest analysis, however, identifies the majority of FSHD1 patients, which is more common than FSHD2. As explained previously, haplotype analysis of 4qA also has to be taken into account in the molecular identification of both FSHD1 and FSHD2, and SMCHD1 deletions in the identification of FSHD2.

The clinical features of both FSDH1 and FSDH2 are similar ( ). Facial and shoulder-girdle weakness are the hallmarks of FSHD, particularly in early stages of the disorder, and may be asymmetrical. Facial weakness is detected by the inability to bury the eyelashes, or in whistling, pursing the lips or puffing out the cheeks. Affected individuals often have a characteristic dimple at the angle of the slightly pouting mouth, which appears when attempting to smile. Shoulder-girdle weakness results in difficulty in raising the arms and, when attempted, a characteristic upward drift of the scapulae is seen. Progression is variable and may be slow and minimal in some. In others, progression may lead to involvement of the trunk and pelvic muscles, resulting in marked lordosis and loss of ambulation. Weakness of abdominal muscles is also frequently seen in later stages, resulting in a protruding abdomen. Hearing loss and a retinal vasculopathy are also common ( ), and only a few cases show cardiorespiratory involvement ( ).

Histology and Histochemistry

With the advent of a reliable molecular diagnostic test for FSHD, muscle biopsies are now rarely performed, but we have retained a description of the pathological features based on historical knowledge. No specific pathological features have been identified in affected patients, despite extensive studies ( ). The degree of pathological changes is variable, which may in some cases relate to the site of biopsy ( Fig. 14.1 ). As discussed in Chapter 1 , it is important to take a sample from an affected muscle but not one that is so severely affected that there are few muscle fibres to examine. Thus, some biopsies from quadriceps muscle in cases where these muscles are relatively less affected may show only very mild changes. Muscles with marked clinical weakness, however, may also show very little pathology. This is another example where clinical severity does not correlate with the degree of pathology.

Increased variability in the size of fibres, with large and small fibres, is common, and the mean fibre diameter of both fibre types is often increased. Very small fibres scattered among larger fibres are a characteristic feature (see Fig. 14.1a ). These have often been described as atrophic fibres, but the presence of developmentally regulated proteins such as fetal myosin, major histocompatibility complex class I (MHC-1I) and desmin raises other possibilities for their small size, such as regeneration ( Fig. 14.2 ). As in Becker muscular dystrophy, clusters of fibres have been put forward as evidence of denervation, but there are no morphological or electrophysiological data to support this. These clusters of small fibres also show the presence of proteins associated with immaturity, including fetal myosin, suggesting that they may represent regeneration, but the presence of fetal myosin does not totally exclude denervation or re-expression (see Ch. 6 ). Neither large-group atrophy nor fibre type grouping is a feature of FSHD, but type 2 fibre predominance may occur, in contrast to type 1 predominance, which is common in other muscular dystrophies. Internal nuclei may occasionally be numerous (see Fig. 14.1b ), and an increase in fibrous and adipose tissue may also occur (see Fig. 14.1b ). Necrosis is rarely seen, but moth-eaten and whorled fibres may occur. An inflammatory response, which may vary from mild to profuse (see Fig. 14.1c ), is common.

Myotonic Dystrophies

Myotonic dystrophy is also an autosomal dominant disorder characterized by myotonia in association with muscle weakness and wasting and also affecting several other tissues. One rare homozygote, however, has been reported, but the homozygosity did not seem to influence the disease phenotype ( ). Two forms of myotonic dystrophy (DM1 and DM2), caused by defects in two different genes, have been identified. DM2 is also known as proximal myotonic myopathy (PROMM). Both are caused by expansion of a nucleotide repeat: DM1 by expansion of a CTG repeat in the 3′ untranslated region of the gene DMPK on chromosome 19q, and DM2 by a CCTG repeat expansion in the first intron of the CNBP gene on chromosome 3q encoding for the cellular nucleic acid-binding protein. The chromosome 19 protein is a putative kinase (DM protein kinase, DMPK) and that from the chromosome 3 gene is a zinc finger protein. Both disorders are believed to result from ‘toxic RNA’ produced by the expansion, which results in the binding of proteins such as muscleblind-like 1 (MBNL1) that leads to abnormal alternative splicing of several pre-mRNAs ( ).

DM1 is a common disorder, with an estimated prevalence of about 1:8000 in Caucasians. DM2, in contrast, is considered more rare, but DM2 is common in some populations and there may be a higher prevalence than is currently appreciated ( ). There are no reported congenital cases of DM2, but congenital presentation of DM1 is well recognized, with the mother almost invariably carrying the mutant gene, and often only mildly affected. Pathology has an important role in differential diagnosis in these cases (see below).

In normal individuals there are between 4 and 40 CTG repeats in the DMPK gene, but in DM1 patients this is increased to 50 or more, sometimes to over 1000. In general, there is a correlation between the size of the repeat and clinical severity and age of onset. Thus those with fewer than 100 repeats are usually milder than the severely affected congenital cases with more than 1000 repeats. Anticipation in DM1 is common, with successive generations being more clinically severe and having more repeats. There is also somatic variability and instability in the size of the repeat expansion, with the number being greater and more unstable in muscle than in blood lymphocytes.

In the normal population the number of chromosome 3 CCTG repeats ranges from 10 to 30, but in DM2 patients this is increased to many thousands. Anticipation and somatic variability also appear to occur in DM2, but the correlations are less clear than in DM1, as fewer affected families have been identified.

Myotonia is common to both DM1 and DM2, but the pattern of muscle weakness is different. In DM2 there is early proximal muscle involvement, in contrast to the distal pattern seen in DM1 ( ): hence the name of PROMM commonly attributed to the disorder. Facial weakness is also rare, or less severe, in DM2 but is common in DM1. Ptosis and facial and neck weakness are characteristic features of DM1. Similarly, cardiac malfunction and central nervous system involvement are less common in DM2 than in DM1. Diaphragm weakness leads to respiratory insufficiency and is often a cause of death in DM1. Cataracts occur in both DM1 and DM2, and other associated features of both include frontal balding, diabetes and gonadal atrophy. Serum creatine kinase (CK) levels are mild to moderately elevated in both ( ).

Molecular analysis of the myotonic dystrophies to confirm a clinical diagnosis is highly reliable, and muscle biopsies are now less often performed. The diagnosis of patients with DM2, however, may not be as clinically apparent as for DM1, and a muscle biopsy from a DM2 patient is then more likely to be taken. Histopathological information on DM2 has increased, as more molecularly confirmed cases have been studied and differences between DM1 and DM2 are apparent, summarized in Table 14.1 ( ).